nickva commented on issue #2430: Running CouchDB in container and then the couchdb pod doesn't release RAM consumed URL: https://github.com/apache/couchdb/issues/2430#issuecomment-584196840 Those message might indicate that the connections between nodes get interrupted. Make sure you have a stable and relatively low latency connection between your cluster nodes. For additional debugging could check out these ares: 1) Does `http $DB/_membership` return the expected set of nodes under all nodes and cluster nodes fields: An example on my machine: ``` $ http $DB/_membership { "all_nodes": [ "[email protected]", "[email protected]", "[email protected]" ], "cluster_nodes": [ "[email protected]", "[email protected]", "[email protected]" ] } ``` 2) Then if it does inspect _stats/rexi and _stats/fabric on some of the nodes: ``` http $DB/_node/_local/_stats/fabric { "doc_update": { "errors": { "desc": "number of document update errors", "type": "counter", "value": 0 }, "mismatched_errors": { "desc": "number of document update errors with multiple error types", "type": "counter", "value": 0 }, "write_quorum_errors": { "desc": "number of write quorum errors", "type": "counter", "value": 0 } }, "open_shard": { "timeouts": { "desc": "number of open shard timeouts", "type": "counter", "value": 0 } }, "read_repairs": { "failure": { "desc": "number of failed read repair operations", "type": "counter", "value": 0 }, "success": { "desc": "number of successful read repair operations", "type": "counter", "value": 0 } }, "worker": { "timeouts": { "desc": "number of worker timeouts", "type": "counter", "value": 0 } } } ``` ``` http $DB/_node/_local/_stats/rexi { "buffered": { "desc": "number of rexi messages buffered", "type": "counter", "value": 0 }, "down": { "desc": "number of rexi_DOWN messages handled", "type": "counter", "value": 0 }, "dropped": { "desc": "number of rexi messages dropped from buffers", "type": "counter", "value": 0 }, "streams": { "timeout": { "init_stream": { "desc": "number of rexi stream initialization timeouts", "type": "counter", "value": 0 }, "stream": { "desc": "number of rexi stream timeouts", "type": "counter", "value": 0 }, "wait_for_ack": { "desc": "number of rexi stream timeouts while waiting for acks", "type": "counter", "value": 0 } } } } ``` 3) If you have remsh installed (in later releases it is part of the release in bin/ folder) you can run it see if connect you to the Erlang VM on one of the nodes and run `mem3:compare_nodelists()` ``` $ remsh Erlang/OTP 21 [erts-10.3.5.3] [source] [64-bit] [smp:12:12] [ds:12:12:10] [async-threads:1] [hipe] Eshell V10.3.5.3 (abort with ^G) ([email protected])1> mem3:compare_nodelists(). [{non_member_nodes,[]}, {bad_nodes,[]}, {{cluster_nodes,['[email protected]','[email protected]', '[email protected]']}, ['[email protected]','[email protected]','[email protected]']}] ```
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
